ExpertsFMEP 150: Bret Taylor (CEO, Sierra): A New Class of Software Winners
The Logan Bartlett Show · September 12, 2025
5 Selected Lessons:
case · 1 framework · 1 heuristic · 1 reasoning_pattern · 1 priority · 1
Business model transition is harder than technology transition
Source
The Logan Bartlett Show — EP 150: Bret Taylor (CEO, Sierra): A New Class of Software Winners (September 12, 2025) · at 19:22 ▸
When this applies
when any organization must change not just its product but its revenue/business model to survive a shift
The takeaway
Rank the difficulty of a transformation by its people/stakeholder-management complexity, not its technical complexity — the hardest part of disruption is usually the business model and change management, not the technology.
Why it works
Technology problems are solvable with engineering effort and improve predictably over time, but business model transitions require simultaneously managing investors' quarterly expectations, employees' incentive structures, and customers' migration — multiple human constituencies who each have reasons to resist change.
One level up
When comparing problems, the socio-organizational dimension (aligning investors, employees, customers through a transition) is generally the binding constraint, not the technical one.
Where it stops holding
This stops working when the technology itself is the genuine bottleneck (e.g. the underlying capability doesn't exist yet), in which case technical difficulty dominates.
Detail
hierarchy:
- Business model/change management difficulty (investor, employee, customer alignment)
- Technology difficulty (building the product)
rationale: Taylor states explicitly technology is hard but business model transitions are harder — you have to be a great storyteller and bring employees, investors, and customers along, and the accounting and reported numbers shift in ways that make a healthy transition look like decline to public market investors.
In their words
so how you actually build a product that's useful for stakeholders one thing but then what is the business model and the the reason why it's hard to make a business model shift is just
Agents shift software valuation from productivity metrics to outcome economics
Source
The Logan Bartlett Show — EP 150: Bret Taylor (CEO, Sierra): A New Class of Software Winners (September 12, 2025) · at 2:30 ▸
When this applies
when a product moves from merely enhancing human output to directly performing the outcome itself
The takeaway
If a product can be repositioned from 'helping someone do a job' to 'doing the job,' the addressable market and pricing logic shift from tool economics to outcome/labor economics.
Why it works
When software only assists a human, its value is capped by productivity-tool economics (per-seat, low TAM); when it performs the actual labor/outcome, its value is tied to the economic value of that outcome itself (e.g. sales commission, legal fee), which is a much larger pool.
One level up
Reframe what market you're actually in by asking whether you're selling a tool or substituting for the thing the tool used to merely assist.
Where it stops holding
This stops working when the agent cannot yet reliably perform the full outcome and is still just assisting a human, in which case it remains priced like traditional productivity software.
Detail
approach: Ask whether the product performs a previously human job outcome versus enhancing human productivity; price and value it against the economic value of that outcome (e.g. a sales commission) rather than against tool-based comparables.
signals:
- TAM for 'productivity enhancement to X professionals' is small
- TAM for 'doing X's job' (e.g. legal labor, sales) is large
- valuation method shifts to margin/commission on the outcome produced
rationale: The traditional view of addressable markets in software is upended because agents don't just enhance people, they replace the labor and thus can be valued like labor/outcome economics.
In their words
As a consequence, I think how you evaluate the value of a piece of software starts to move away from traditional software productivity metrics. Think about a CIRA agent that actually makes a sale for you. The way you would value that is not really even related to AI or software. The way you'd value it is basically one of the margins on that sale.
Look for second-order behavioral effects when a cost collapses, not just first-order savings
Source
The Logan Bartlett Show — EP 150: Bret Taylor (CEO, Sierra): A New Class of Software Winners (September 12, 2025) · at 11:15 ▸
When this applies
when a new technology drastically lowers the marginal cost of an activity that was previously rationed due to cost
The takeaway
When evaluating a cost-reducing technology, ask not 'how much cost will this save' but 'what will people now do more of, and does that change the fundamental purpose of the activity' — second-order behavioral shifts often dwarf first-order cost savings.
Why it works
When the marginal cost of an action drops by orders of magnitude, businesses don't just recoup the savings — they change their behavior and do vastly more of the action, which can unlock entirely new strategic uses (e.g. turning a cost center into a revenue driver through customer engagement).
One level up
The biggest value from a step-change cost reduction usually comes from category-changing behavior change, not from efficiency gains within the old category.
Where it stops holding
This stops working when the activity's volume is already unconstrained by cost (so cheaper costs don't unlock new behavior) or when the business has no lever to convert increased activity into revenue.
Detail
approach: Identify what a 100x cost reduction in an activity would let an organization do differently in volume or kind, especially in businesses driven by lifetime value or customer relationships.
signals:
- cost of an action approaching the cost of a near-free action (e.g. a phone call approaching the price of a page view)
- business model built on lifetime value/attrition, not just unit cost
- an existing cost center (e.g. call center) that could become a growth lever
rationale: People often miss second order effects; the 'obvious' win (OpEx savings) is dwarfed by the strategic shift of treating the activity as growth infrastructure instead of overhead.
In their words
When the price of a phone call starts to approach the price of a page view, you're going to do a lot more of them. And as a consequence, I think it's just going to really up end to the markets.
Measurability of business value determines pricing power and compression
Source
The Logan Bartlett Show — EP 150: Bret Taylor (CEO, Sierra): A New Class of Software Winners (September 12, 2025) · at 6:03 ▸
When this applies
when pricing any product and deciding whether per-seat/flat pricing or outcome-based pricing captures more value
The takeaway
The more measurable and attributable a product's business outcome is, the closer its price can get to the value of that outcome rather than being compressed by comparison to substitute tools.
Why it works
Horizontal tools deliver value that is hard to measure per user (same price for a junior employee and a senior R&D engineer), which commoditizes price; departmental/outcome-oriented tools deliver measurable business value (a closed sale, an audited ledger), allowing pricing closer to that value, often an order of magnitude higher per seat.
One level up
Pricing power comes from proximity to measurable value, not from technology superiority — the same logic that moved digital advertising from impressions to cost-per-click/conversion.
Where it stops holding
This stops working when the outcome is genuinely hard to isolate and attribute to the tool (e.g. value of an individual video call), in which case outcome-based pricing isn't feasible and commoditized per-seat pricing persists.
Detail
rule: Price as close as possible to the measurable business outcome your product drives, not to the cost or feature set of competing technology.
threshold: Order-of-magnitude difference: Taylor states departmental enterprise software (ServiceNow, Salesforce, SAP) captures 'more than an order of magnitude' more value per seat than horizontal tools like Zoom/Slack, despite far fewer users.
rationale: History already played out this way in advertising (CPM to CPC/conversion); the same logic applies to the digital economy broadly — the closer you are to a measurable accountable outcome, the more value accrues to your platform.
In their words
I think in the sort of digital economy, the closest you can get to a measurable accountable outcome, the more value will accruate your platform.
AltaVista vs Google vs Inktomi — business model beats technical quality
Source
The Logan Bartlett Show — EP 150: Bret Taylor (CEO, Sierra): A New Class of Software Winners (September 12, 2025) · at 0:00 ▸
When this applies
when multiple competitors chase an obviously valuable new market and the winner is unclear
The takeaway
In an obviously valuable new market, winners are decided less by who builds the best technology and more by who picks the business model that lets them capture the value they create.
Why it works
Technical superiority alone does not determine market winner; the business model determines whether a company can capture the full value the technology makes possible.
One level up
When a market opportunity is self-evident, the differentiator shifts from 'does this work' to 'how do we monetize it' — competitive advantage moves downstream to execution and packaging.
Where it stops holding
This stops working when the market or opportunity itself is not yet obvious, so insight into the opportunity itself (not just the model) is the scarce resource.
Detail
situation: Multiple companies (AltaVista, Google, Inktomi) competed in the obviously valuable search market with different technical quality and business models.
decision: Inktomi, despite strong engineering, licensed its search engine to portals (B2B) rather than building a consumer-facing ad business; Google built its own consumer product and ad auction.
outcome: Google's B2C model and ad business became, in Taylor's words, 'the greatest business model of all time,' while Inktomi and AltaVista faded.
lesson: The concept being obvious does not mean the winner is obvious; business model and go-to-market choice decide the outcome.
In their words
So there's so many details that, you know, dictate whether you're the one you turn into Google or altavista or ink to me.